Recent studies demonstrate gastrinomas, similar to carcinoid tumors and all other pancreatic endocrine tumors (PET's) except insulinomas, frequently (60-90%) are malignant and some may have an aggressive course. The molecular pathogenesis, not only for the tumors themselves, but also for determining their growth behavior is almost completely unknown. Recent studies demonstrate that in contrast to many other more common cancers neither oncogenes nor common tumor suppressor genes (p53, retinoblastoma, VHL gene, etc.) are generally altered in gastrinomas, carcinoids or other PET's. Recent analyses at NIH have identified a cohort of 25% of patients in whom gastrinomas pursued an aggressive clinical course. Both clinical and laboratory characteristics that distinguish patients with an aggressive course are being sought by statistical analysis as well as correlations with possible molecular changes in the gastrinoma that may correlate with prognosis and tumor growth. Our recent studies demonstrate that the tumor suppressor gene, p16, which is involved in maintaining cell cycle control, is frequently (50%) altered in gastrinomas, entirely due to methylation of CG-rich islands in the promoter region. At present we are investigating the importance of overexpression of growth factors (EGFR, HGFR)in gastrinomas, alterations in the HER-2/neu oncogene, which was reported to occur in a small number of PET's and have undertaken the possible identification of important other molecular alterations in determining tumor growth using genomic-wide allelotyping.